Finding the Predictive Validity Coefficient of a Test

I'm in psychometrics and am historically absolutely horrible at all forms of mathematics post algebra 1.

I have 2 sets of data, one is the scores students have gotten on a new reading test and the second is what their actual grades were.
I need to find the predictive validity coefficient of the test.

ill post the data
x y
62 74
73 93
88 68
82 79
85 91
77 72
94 96
65 61
91 92
74 82
85 93
98 95

were x is the test scores and y is the actual grades.

sum of x, squared is 948,676
sum of x squareds is 80,442
sum of y, squared is 992,016
sum of y squareds is 84,254

mean of x is 81.17 mean of y is 83

So what do I have to do from here I have no clue? I looked through my text book, looked at previous questions I've done, looked on youtube, googled it, etc.

It seems the internet just wants to tell me what predictive validity is and not how to actually get it!

Do I just do a pearson's r on the information? Or is there an equation for this?

I expect you need a "Correlation Coefficient". It is often known as R^2.

I get 0.3851 for your data. Can you reproduce this?

Note:

"I ... am historically absolutely horrible at all forms of mathematics post algebra 1."

Please see a doctor about this and get it fixed before you are entirely incapacitated.

tkhunny said:

I expect you need a "Correlation Coefficient". It is often known as R^2.

I get 0.3851 for your data. Can you reproduce this?

Note:

"I ... am historically absolutely horrible at all forms of mathematics post algebra 1."

Please see a doctor about this and get it fixed before you are entirely incapacitated.

Click to expand...

no I got 0.62

actually if I do an R^2 on my 0.62 I get 0.3844 is that right? I am a few decimal points off?

Which equation did you use?
I only have pearson's r and kuder richardson lying around.

My bad. It's usually

r = correlation coefficient and
r^2 = coefficient of determination

\(\displaystyle r = \frac{n\sum xy - \left(\sum x\right)\left(\sum y\right)}{\sqrt{n\sum x^{2} - \left(\sum x\right)^{2}}\sqrt{n\sum y^{2} - \left(\sum y\right)^{2}}}\)

tkhunny said:

My bad. It's usually

r = correlation coefficient and
r^2 = coefficient of determination

\(\displaystyle r = \frac{n\sum xy - \left(\sum x\right)\left(\sum y\right)}{\sqrt{n\sum x^{2} - \left(\sum x\right)^{2}}\sqrt{n\sum y^{2} - \left(\sum y\right)^{2}}}\)

Click to expand...

hey, oddly enough, I have the same question... May I ask, Tkhunny, what numbers do you input where into that formula? Where did you find this formula, is there a website that'll assist with using this formula?

Thanks